US 8195674 B1 Abstract A system for generating a model is provided. The system generates, or selects, candidate conditions and generates, or otherwise obtains, statistics regarding the candidate conditions. The system also forms rules based, at least in part, on the statistics and the candidate conditions and selectively adds the rules to the model.
Claims(18) 1. A computer-implemented method, comprising:
generating, by one or more server devices, a model that is based on:
a condition that includes a conjunction of a plurality of features associated with documents, and
statistics, received from one or more other devices, associated with the condition, where a particular statistic from a particular device, of the one or more other devices, indicates a weight, determined by the particular device, for the condition,
generating the model including:
generating a new weight for the condition, for use in the model, based on the received statistics from the one or more other devices, the new weight for the condition indicating how relevant the condition is, with respect to other conditions in the model, in determining how relevant a document is to a search query,
generating a candidate rule for the model based on the condition and the received statistics,
determining whether to add the candidate rule to the model, and
upon determining that the candidate rule should not be added to the model, setting a weight, for the rule, to a value that indicates that the candidate rule should not be added to the model; and
using, by the one or more server devices, the model to generate a rank for a document, the rank indicating an estimated probability that the document is relevant to a received search query.
2. The computer-implemented method of
3. The computer-implemented method of
4. The computer-implemented method of
replacing, in the model, a previous rule, that is based on the condition and a previous weight, with a new rule that is based on the condition and the new weight.
5. The computer-implemented method of
selectively replacing the previous rule with the new rule, based on a cost associated with the condition.
6. The computer-implemented method of
generating a new rule that is based on the condition and the new weight, and
adding the new rule to the model.
7. The computer-implemented method of
8. The computer-implemented method of
generating a rule based on the condition and the received statistics, and
sending the rule to the one or more other devices.
9. The computer-implemented method of
10. The computer-implemented method of
generating a candidate rule for the model based on the condition and the received statistics,
determining whether to add the candidate rule to the model, and
upon determining that the candidate rule should not be added to the model, informing the one or more other devices that the candidate rule should not be added to the model.
11. A system comprising:
a memory to store computer-readable instructions; and
one or more processors to execute the instructions to:
generate a model that is based on a condition that includes a conjunction of a plurality of features associated with documents, and statistics, received from one or more other devices, associated with the condition, where a particular statistic from a particular device, of the one or more other devices, indicates a weight, determined by the particular device, for the condition, the processor, when generating of the model being further to:
generate a new weight for the condition, for use in the model, based on the received statistics from the one or more other devices, the new weight for the condition indicating how relevant the condition is, with respect to other conditions in the model, in determining how relevant a document is to a search query, and
generate a candidate rule for the model based on the condition and the received statistics,
determine whether to add the candidate rule to the model, and
upon determining that the candidate rule should not be added to the model, set a weight, for the rule, to a value that indicates that the candidate rule should not be added to the model; and
use the model to generate a rank for a document, the rank indicating an estimated probability that the document is relevant to a received search query.
12. The system of
13. The system of
generate a new rule that is based on the condition and the new weight, and
add the new rule to the model.
14. A memory device comprising:
one or more instructions which, when executed by a computer device, cause the computer device to generate a model that is based on a condition that includes a conjunction of a plurality of features associated with documents, and statistics, received from one or more other devices, associated with the condition, where a particular statistic from a particular device, of the one or more other devices, indicates a weight, determined by the particular device, for the condition,
the one or more instructions to generate the model including:
one or more instructions to generate a new weight for the condition, for use in the model, based on the received statistics from the one or more other devices, the new weight for the condition indicating how relevant the condition is, with respect to other conditions in the model, in determining how relevant a document is to a search query,
one or more instructions to generate a candidate rule for the model based on the condition and the received statistics,
one or more instructions to determine whether to add the candidate rule to the model, and
one or more instructions to set, upon determining that the candidate rule should not be added to the model, a weight, for the rule, to a value that indicates that the candidate rule should not be added to the model; and
one or more instructions which, when executed by the computer device, cause the computer device to use the model to generate a rank for a document, the rank indicating an estimated probability that the document is relevant to a received search query.
15. The memory device of
one or more instructions to replace, in the model, a previous rule, that is based on the condition and a previous weight, with a new rule that is based on the condition and the new weight, and
where the one or more instructions to replace the previous rule include:
one or more instructions to selectively replace the previous rule with the new rule, based on a cost associated with the condition.
16. The memory device of
one or more instructions to generate a new rule that is based on the condition and the new weight; and
one or more instructions to add the new rule to the mode, where the one or more instructions to add the new rule includes:
one or more instructions to add the new rule selectively, based on a cost associated with the condition.
17. The memory device of
one or more instructions to generate a rule based on the condition and the received statistics; and
one or more instructions to send the rule to the one or more other devices.
18. The memory device of
one or more instructions to generate a candidate rule for the model based on the condition and the received statistics;
one or more instructions to determine whether to add the candidate rule to the model; and
one or more instructions to inform, upon determining that the candidate rule should not be added to the model, the one or more other devices that the candidate rule should not be added to the model.
Description This application is a continuation of U.S. patent application Ser. No. 11/736,193, filed Apr. 17, 2007, which is a continuation of U.S. patent application Ser. No. 10/734,584, filed Dec. 15, 2003 (now U.S. Pat. No. 7,222,127), which is a continuation-in-part of U.S. patent application Ser. No. 10/706,991, filed Nov. 14, 2003 (now U.S. Pat. No. 7,231,399), the disclosures of which are incorporated herein by reference. 1. Field of the Invention The present invention relates generally to classification systems and, more particularly, to systems and methods for applying machine learning to various large data sets to generate a classification model. 2. Description of Related Art Classification models have been used to classify a variety of elements. The classification models are built from a set of training data that usually includes examples or records, each having multiple attributes or features. The objective of classification is to analyze the training data and develop an accurate model using the features present in the training data. The model is then used to classify future data for which the classification is unknown. Several classification systems have been proposed over the years, including systems based on neural networks, statistical models, decision trees, and genetic models. One problem associated with existing classification systems has to do with the volume of training data that they are capable of handling. Existing classification systems can only efficiently handle small quantities of training data. They struggle to deal with large quantities of data, such as more than one hundred thousand features. Accordingly, there is a need for systems and methods that are capable of generating a classification model from a large data set. Systems and methods, consistent with the principles of the invention, apply machine learning to large data sets to generate a classification model. In accordance with one aspect consistent with the principles of the invention, a system for generating a model is provided. The system may include multiple nodes. At least one of the nodes is configured to select a candidate condition, request statistics associated with the candidate condition from other ones of the nodes, receive the requested statistics from the other nodes, form a rule based, at least in part, on the candidate condition and the requested statistics, and selectively add the rule to the model. According to another aspect, a system for generating a model is provided. The system may form candidate conditions and generate statistics associated with the candidate conditions. The system may also form rules based, at least in part, on the candidate conditions and the generated statistics and selectively add the rules to the model. According to yet another aspect, a method for generating a model in a system that includes multiple nodes is provided. The method may include generating candidate conditions, distributing the candidate conditions to the nodes, and generating statistics regarding the candidate conditions. The method may also include collecting the statistics for each of the candidate conditions at one of the nodes, generating rules based, at least in part, on the statistics and the candidate conditions, and selectively adding the rules to the model. According to a further aspect, a system for generating a model is provided. The system may generate new conditions and distribute the new conditions to a set of nodes. Each of the nodes may generate statistics regarding the new conditions. The system may generate new rules based, at least in part, on the statistics and the new conditions and add at least one of the new rules to the model. The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate an embodiment of the invention and, together with the description, explain the invention. In the drawings, The following detailed description of the invention refers to the accompanying drawings. The same reference numbers in different drawings may identify the same or similar elements. Also, the following detailed description does not limit the invention. Systems and methods consistent with the principles of the invention may apply machine learning to large data sets, such as data sets including over one hundred thousand features and/or one million instances. The systems and methods may be capable of processing a large data set in a reasonable amount of time to generate a classification model. Different models may be generated for use in different contexts. For example, in an exemplary e-mail context, a model may be generated to classify e-mail as either spam or normal (non-spam) e-mail. In an exemplary advertisement context, a model may be generated to estimate the probability that a user will click on a particular advertisement. In an exemplary document ranking context, a model may be generated in connection with a search to estimate the probability that a user will find a particular search result relevant. Other models may be generated in other contexts where a large number of data items exist as training data to train the model. Repository In the exemplary e-mail context, the data set in repository A feature X may be an aspect of the domain (e.g., the e-mail domain) that may be useful to determine the label (e.g., “the number of exclamation points in the message” or “whether the word ‘free’ appears in the message”). In one implementation, each feature X may include a boolean value (e.g., a value of zero or one based on whether the word “free” appears in the message). In another implementation, each feature X may include a discrete value (e.g., a value based, at least in part, on the number of exclamation points in the message). In yet another implementation, each feature X may include a real value (e.g., the time of day a message was sent). An instance d may be written as: d=(x Repository Nodes Each of nodes Processor Input device(s) As will be described in detail below, node The software instructions may be read into memory To facilitate generation of the model, a prior probability of the label for each instance may be determined: P(Y|Z). This prior probability can be based on Z, which may include one or more values that differ based on the particular context in which the model is used. Typically, Z may be real valued and dense (i.e., it does not include many zero entries for many of the instances). In the e-mail context, each e-mail may be evaluated using a common spam detection program that gives each e-mail a score (e.g., Spam Assassin). The output of the spam detection program may be used as the prior probability that the e-mail is spam. A set of instances based on the same or a different set of instances as in repository As will be explained later, it may be beneficial to quickly obtain statistics for the instances that contain particular features. To facilitate fast identification of correspondence between features and instances, a feature-to-instance index may be generated in some implementations to link features to the instances in which they are included. For example, for a given feature X, the set of instances that contain that feature may be listed. The list of instances for a feature X is called the “hitlist for feature X.” Thereafter, given a set of features X A “condition” C is a conjunction of features and possibly their complements. For example, a condition that includes two features is: “the message contains the word ‘free’” and “the domain of the sender is “hotmail.com,” and a condition that includes a feature and a complement of a feature is: “the message contains the word ‘free’” and “the domain of the sender is not ‘netscape.net.’” For any instance d Based, at least in part, on this information, a function may be created that maps conditions to a probability of the label: P(Y|C Thereafter, given a new instance d and a model M, the posterior probability of the label may be determined by: (1) extracting the features from the instance, (2) determining which rules apply, and (3) combining the weight of each rule with the prior probability for instance d. Therefore, the goal is to generate a good model. To generate a good model, the following information may be beneficial: the set of conditions C Processing may begin with an empty model M that includes the prior probability of the label. A node Node Node Each of nodes Node Node Node Processing may then return to act As described previously, the acts described with respect to Processing may begin with an empty model M that includes the prior probability of the label. A node Node Each of nodes Node Node Node Processing may then return to act As described previously, the acts described with respect to Generally, the processing of Processing may begin with the generation of new conditions as candidate conditions to test whether they would make good rules for the model M (act The goal of the candidate rule generation phase is to generate new conditions that match some minimum number of instances. There are a couple of strategies for accomplishing this. For example, conditions that appear multiple times in some fraction of the instances (divided among all of nodes Alternatively, conditions that appear a certain number of times on a single node Then in the rule testing and optimization phase, the candidate conditions may be distributed to all nodes Each node Each node Node The rule testing and optimization phase may continue for a number of iterations or until all rules have been tested. The output of the rule testing and optimization phase is new weights for all existing rules (possibly zero if the rule is to be dropped from the model M) and a list of new rules. As described previously, the acts described with respect to Systems and methods consistent with the principles of the invention may generate a model from a large data set (e.g., a data set that includes possibly millions of data items) efficiently on multiple nodes. The foregoing description of preferred embodiments of the present invention provides illustration and description, but is not intended to be exhaustive or to limit the invention to the precise form disclosed. Modifications and variations are possible in light of the above teachings or may be acquired from practice of the invention. For example, while series of acts have been described with regard to Also, in the three implementations described with regard to It will also be apparent to one of ordinary skill in the art that aspects of the invention, as described above, may be implemented in many different forms of software, firmware, and hardware in the implementations illustrated in the figures. The actual software code or specialized control hardware used to implement aspects consistent with the present invention is not limiting of the present invention. Thus, the operation and behavior of the aspects were described without reference to the specific software code—it being understood that one of ordinary skill in the art would be able to design software and control hardware to implement the aspects based on the description herein. Patent Citations
Non-Patent Citations
Referenced by
Classifications
Rotate |